preprocess.py 13.4 KB
Newer Older
Louis Martin's avatar
Louis Martin committed
1
#!/usr/bin/env python3
2
# Copyright (c) Facebook, Inc. and its affiliates.
Sergey Edunov's avatar
Sergey Edunov committed
3
#
4
5
# This source code is licensed under the MIT license found in the
# LICENSE file in the root directory of this source tree.
Myle Ott's avatar
Myle Ott committed
6
7
8
"""
Data pre-processing: build vocabularies and binarize training data.
"""
Sergey Edunov's avatar
Sergey Edunov committed
9

Sergey Edunov's avatar
Sergey Edunov committed
10
from collections import Counter
Sergey Edunov's avatar
Sergey Edunov committed
11
12
from itertools import zip_longest

Myle Ott's avatar
Myle Ott committed
13
from fairseq import options, tasks, utils
14
from fairseq.data import indexed_dataset
15
from fairseq.binarizer import Binarizer
Myle Ott's avatar
Myle Ott committed
16
from multiprocessing import Pool
Sergey Edunov's avatar
Sergey Edunov committed
17

18
19
import os
import shutil
20

Sergey Edunov's avatar
Sergey Edunov committed
21

Myle Ott's avatar
Myle Ott committed
22
def main(args):
Myle Ott's avatar
Myle Ott committed
23
    utils.import_user_module(args)
24

Sergey Edunov's avatar
Sergey Edunov committed
25
    print(args)
26

Sergey Edunov's avatar
Sergey Edunov committed
27
    os.makedirs(args.destdir, exist_ok=True)
28
    target = not args.only_source
Sergey Edunov's avatar
Sergey Edunov committed
29

30
31
    task = tasks.get_task(args.task)

alexeib's avatar
alexeib committed
32
    def train_path(lang):
Ruty Rinott's avatar
Ruty Rinott committed
33
        return "{}{}".format(args.trainpref, ("." + lang) if lang else "")
alexeib's avatar
alexeib committed
34
35
36
37

    def file_name(prefix, lang):
        fname = prefix
        if lang is not None:
Myle Ott's avatar
Myle Ott committed
38
            fname += ".{lang}".format(lang=lang)
alexeib's avatar
alexeib committed
39
40
41
42
43
44
        return fname

    def dest_path(prefix, lang):
        return os.path.join(args.destdir, file_name(prefix, lang))

    def dict_path(lang):
Ruty Rinott's avatar
Ruty Rinott committed
45
        return dest_path("dict", lang) + ".txt"
alexeib's avatar
alexeib committed
46

47
48
49
50
51
52
53
54
    def build_dictionary(filenames, src=False, tgt=False):
        assert src ^ tgt
        return task.build_dictionary(
            filenames,
            workers=args.workers,
            threshold=args.thresholdsrc if src else args.thresholdtgt,
            nwords=args.nwordssrc if src else args.nwordstgt,
            padding_factor=args.padding_factor,
Ruty Rinott's avatar
Ruty Rinott committed
55
        )
56

57
58
59
60
61
    if not args.srcdict and os.path.exists(dict_path(args.source_lang)):
        raise FileExistsError(dict_path(args.source_lang))
    if target and not args.tgtdict and os.path.exists(dict_path(args.target_lang)):
        raise FileExistsError(dict_path(args.target_lang))

62
    if args.joined_dictionary:
63
64
        assert not args.srcdict or not args.tgtdict, \
            "cannot use both --srcdict and --tgtdict with --joined-dictionary"
65
66
67
68
69
70

        if args.srcdict:
            src_dict = task.load_dictionary(args.srcdict)
        elif args.tgtdict:
            src_dict = task.load_dictionary(args.tgtdict)
        else:
71
72
73
74
            assert args.trainpref, "--trainpref must be set if --srcdict is not specified"
            src_dict = build_dictionary(
                {train_path(lang) for lang in [args.source_lang, args.target_lang]}, src=True
            )
Myle Ott's avatar
Myle Ott committed
75
        tgt_dict = src_dict
76
    else:
Myle Ott's avatar
Myle Ott committed
77
        if args.srcdict:
78
            src_dict = task.load_dictionary(args.srcdict)
Myle Ott's avatar
Myle Ott committed
79
        else:
80
            assert args.trainpref, "--trainpref must be set if --srcdict is not specified"
81
82
            src_dict = build_dictionary([train_path(args.source_lang)], src=True)

83
84
        if target:
            if args.tgtdict:
85
                tgt_dict = task.load_dictionary(args.tgtdict)
86
            else:
87
                assert args.trainpref, "--trainpref must be set if --tgtdict is not specified"
88
89
90
                tgt_dict = build_dictionary([train_path(args.target_lang)], tgt=True)
        else:
            tgt_dict = None
Myle Ott's avatar
Myle Ott committed
91

alexeib's avatar
alexeib committed
92
    src_dict.save(dict_path(args.source_lang))
93
    if target and tgt_dict is not None:
alexeib's avatar
alexeib committed
94
        tgt_dict.save(dict_path(args.target_lang))
Sergey Edunov's avatar
Sergey Edunov committed
95

96
97
    def make_binary_dataset(vocab, input_prefix, output_prefix, lang, num_workers):
        print("| [{}] Dictionary: {} types".format(lang, len(vocab) - 1))
Sergey Edunov's avatar
Sergey Edunov committed
98
99
        n_seq_tok = [0, 0]
        replaced = Counter()
Sergey Edunov's avatar
Sergey Edunov committed
100

Sergey Edunov's avatar
Sergey Edunov committed
101
        def merge_result(worker_result):
Ruty Rinott's avatar
Ruty Rinott committed
102
103
104
            replaced.update(worker_result["replaced"])
            n_seq_tok[0] += worker_result["nseq"]
            n_seq_tok[1] += worker_result["ntok"]
Sergey Edunov's avatar
Sergey Edunov committed
105

Ruty Rinott's avatar
Ruty Rinott committed
106
107
108
        input_file = "{}{}".format(
            input_prefix, ("." + lang) if lang is not None else ""
        )
109
        offsets = Binarizer.find_offsets(input_file, num_workers)
Sergey Edunov's avatar
Sergey Edunov committed
110
111
        pool = None
        if num_workers > 1:
Ruty Rinott's avatar
Ruty Rinott committed
112
            pool = Pool(processes=num_workers - 1)
Sergey Edunov's avatar
Sergey Edunov committed
113
114
            for worker_id in range(1, num_workers):
                prefix = "{}{}".format(output_prefix, worker_id)
Ruty Rinott's avatar
Ruty Rinott committed
115
116
117
118
119
                pool.apply_async(
                    binarize,
                    (
                        args,
                        input_file,
120
                        vocab,
Ruty Rinott's avatar
Ruty Rinott committed
121
122
123
                        prefix,
                        lang,
                        offsets[worker_id],
124
                        offsets[worker_id + 1]
Ruty Rinott's avatar
Ruty Rinott committed
125
                    ),
126
                    callback=merge_result
Ruty Rinott's avatar
Ruty Rinott committed
127
                )
Sergey Edunov's avatar
Sergey Edunov committed
128
129
            pool.close()

130
131
        ds = indexed_dataset.make_builder(dataset_dest_file(args, output_prefix, lang, "bin"),
                                          impl=args.dataset_impl, vocab_size=len(vocab))
Ruty Rinott's avatar
Ruty Rinott committed
132
        merge_result(
133
134
135
            Binarizer.binarize(
                input_file, vocab, lambda t: ds.add_item(t),
                offset=0, end=offsets[1]
Ruty Rinott's avatar
Ruty Rinott committed
136
137
            )
        )
Sergey Edunov's avatar
Sergey Edunov committed
138
139
140
141
142
143
144
145
146
        if num_workers > 1:
            pool.join()
            for worker_id in range(1, num_workers):
                prefix = "{}{}".format(output_prefix, worker_id)
                temp_file_path = dataset_dest_prefix(args, prefix, lang)
                ds.merge_file_(temp_file_path)
                os.remove(indexed_dataset.data_file_path(temp_file_path))
                os.remove(indexed_dataset.index_file_path(temp_file_path))

Ruty Rinott's avatar
Ruty Rinott committed
147
        ds.finalize(dataset_dest_file(args, output_prefix, lang, "idx"))
Sergey Edunov's avatar
Sergey Edunov committed
148

Ruty Rinott's avatar
Ruty Rinott committed
149
150
151
152
153
154
155
        print(
            "| [{}] {}: {} sents, {} tokens, {:.3}% replaced by {}".format(
                lang,
                input_file,
                n_seq_tok[0],
                n_seq_tok[1],
                100 * sum(replaced.values()) / n_seq_tok[1],
156
                vocab.unk_word,
Ruty Rinott's avatar
Ruty Rinott committed
157
158
            )
        )
Sergey Edunov's avatar
Sergey Edunov committed
159

160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
    def make_binary_alignment_dataset(input_prefix, output_prefix, num_workers):
        nseq = [0]

        def merge_result(worker_result):
            nseq[0] += worker_result['nseq']

        input_file = input_prefix
        offsets = Binarizer.find_offsets(input_file, num_workers)
        pool = None
        if num_workers > 1:
            pool = Pool(processes=num_workers - 1)
            for worker_id in range(1, num_workers):
                prefix = "{}{}".format(output_prefix, worker_id)
                pool.apply_async(
                    binarize_alignments,
                    (
                        args,
                        input_file,
                        utils.parse_alignment,
                        prefix,
                        offsets[worker_id],
                        offsets[worker_id + 1]
                    ),
                    callback=merge_result
                )
            pool.close()

        ds = indexed_dataset.make_builder(dataset_dest_file(args, output_prefix, None, "bin"),
                                          impl=args.dataset_impl)

        merge_result(
            Binarizer.binarize_alignments(
                input_file, utils.parse_alignment, lambda t: ds.add_item(t),
                offset=0, end=offsets[1]
            )
        )
        if num_workers > 1:
            pool.join()
            for worker_id in range(1, num_workers):
                prefix = "{}{}".format(output_prefix, worker_id)
                temp_file_path = dataset_dest_prefix(args, prefix, None)
                ds.merge_file_(temp_file_path)
                os.remove(indexed_dataset.data_file_path(temp_file_path))
                os.remove(indexed_dataset.index_file_path(temp_file_path))

        ds.finalize(dataset_dest_file(args, output_prefix, None, "idx"))

        print(
            "| [alignments] {}: parsed {} alignments".format(
                input_file,
                nseq[0]
            )
        )

214
    def make_dataset(vocab, input_prefix, output_prefix, lang, num_workers=1):
215
        if args.dataset_impl == "raw":
216
            # Copy original text file to destination folder
217
            output_text_file = dest_path(
Ruty Rinott's avatar
Ruty Rinott committed
218
                output_prefix + ".{}-{}".format(args.source_lang, args.target_lang),
219
220
                lang,
            )
alexeib's avatar
alexeib committed
221
            shutil.copyfile(file_name(input_prefix, lang), output_text_file)
222
223
        else:
            make_binary_dataset(vocab, input_prefix, output_prefix, lang, num_workers)
224

225
    def make_all(lang, vocab):
226
        if args.trainpref:
227
            make_dataset(vocab, args.trainpref, "train", lang, num_workers=args.workers)
228
        if args.validpref:
Ruty Rinott's avatar
Ruty Rinott committed
229
230
            for k, validpref in enumerate(args.validpref.split(",")):
                outprefix = "valid{}".format(k) if k > 0 else "valid"
231
                make_dataset(vocab, validpref, outprefix, lang, num_workers=args.workers)
232
        if args.testpref:
Ruty Rinott's avatar
Ruty Rinott committed
233
234
            for k, testpref in enumerate(args.testpref.split(",")):
                outprefix = "test{}".format(k) if k > 0 else "test"
235
                make_dataset(vocab, testpref, outprefix, lang, num_workers=args.workers)
236

237
238
239
240
241
242
243
244
    def make_all_alignments():
        if args.trainpref and os.path.exists(args.trainpref + "." + args.align_suffix):
            make_binary_alignment_dataset(args.trainpref + "." + args.align_suffix, "train.align", num_workers=args.workers)
        if args.validpref and os.path.exists(args.validpref + "." + args.align_suffix):
            make_binary_alignment_dataset(args.validpref + "." + args.align_suffix, "valid.align", num_workers=args.workers)
        if args.testpref and os.path.exists(args.testpref + "." + args.align_suffix):
            make_binary_alignment_dataset(args.testpref + "." + args.align_suffix, "test.align", num_workers=args.workers)

245
    make_all(args.source_lang, src_dict)
246
    if target:
247
        make_all(args.target_lang, tgt_dict)
248
249
    if args.align_suffix:
        make_all_alignments()
250

Ruty Rinott's avatar
Ruty Rinott committed
251
    print("| Wrote preprocessed data to {}".format(args.destdir))
Sergey Edunov's avatar
Sergey Edunov committed
252
253

    if args.alignfile:
254
        assert args.trainpref, "--trainpref must be set if --alignfile is specified"
alexeib's avatar
alexeib committed
255
256
        src_file_name = train_path(args.source_lang)
        tgt_file_name = train_path(args.target_lang)
Sergey Edunov's avatar
Sergey Edunov committed
257
        freq_map = {}
258
259
260
        with open(args.alignfile, "r", encoding='utf-8') as align_file:
            with open(src_file_name, "r", encoding='utf-8') as src_file:
                with open(tgt_file_name, "r", encoding='utf-8') as tgt_file:
Sergey Edunov's avatar
Sergey Edunov committed
261
                    for a, s, t in zip_longest(align_file, src_file, tgt_file):
262
263
                        si = src_dict.encode_line(s, add_if_not_exist=False)
                        ti = tgt_dict.encode_line(t, add_if_not_exist=False)
Ruty Rinott's avatar
Ruty Rinott committed
264
                        ai = list(map(lambda x: tuple(x.split("-")), a.split()))
Sergey Edunov's avatar
Sergey Edunov committed
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
                        for sai, tai in ai:
                            srcidx = si[int(sai)]
                            tgtidx = ti[int(tai)]
                            if srcidx != src_dict.unk() and tgtidx != tgt_dict.unk():
                                assert srcidx != src_dict.pad()
                                assert srcidx != src_dict.eos()
                                assert tgtidx != tgt_dict.pad()
                                assert tgtidx != tgt_dict.eos()

                                if srcidx not in freq_map:
                                    freq_map[srcidx] = {}
                                if tgtidx not in freq_map[srcidx]:
                                    freq_map[srcidx][tgtidx] = 1
                                else:
                                    freq_map[srcidx][tgtidx] += 1

        align_dict = {}
        for srcidx in freq_map.keys():
            align_dict[srcidx] = max(freq_map[srcidx], key=freq_map[srcidx].get)

Ruty Rinott's avatar
Ruty Rinott committed
285
        with open(
286
287
288
289
290
                os.path.join(
                    args.destdir,
                    "alignment.{}-{}.txt".format(args.source_lang, args.target_lang),
                ),
                "w", encoding='utf-8'
Ruty Rinott's avatar
Ruty Rinott committed
291
        ) as f:
Sergey Edunov's avatar
Sergey Edunov committed
292
            for k, v in align_dict.items():
Ruty Rinott's avatar
Ruty Rinott committed
293
294
295
                print("{} {}".format(src_dict[k], tgt_dict[v]), file=f)


296
def binarize(args, filename, vocab, output_prefix, lang, offset, end, append_eos=True):
297
298
    ds = indexed_dataset.make_builder(dataset_dest_file(args, output_prefix, lang, "bin"),
                                      impl=args.dataset_impl, vocab_size=len(vocab))
Sergey Edunov's avatar
Sergey Edunov committed
299
300
301
302

    def consumer(tensor):
        ds.add_item(tensor)

303
304
    res = Binarizer.binarize(filename, vocab, consumer, append_eos=append_eos,
                             offset=offset, end=end)
Ruty Rinott's avatar
Ruty Rinott committed
305
    ds.finalize(dataset_dest_file(args, output_prefix, lang, "idx"))
Sergey Edunov's avatar
Sergey Edunov committed
306
307
    return res

Ruty Rinott's avatar
Ruty Rinott committed
308

309
310
311
312
313
314
315
316
317
318
319
320
321
def binarize_alignments(args, filename, parse_alignment, output_prefix, offset, end):
    ds = indexed_dataset.make_builder(dataset_dest_file(args, output_prefix, None, "bin"),
                                      impl=args.dataset_impl, vocab_size=None)

    def consumer(tensor):
        ds.add_item(tensor)

    res = Binarizer.binarize_alignments(filename, parse_alignment, consumer, offset=offset,
                                        end=end)
    ds.finalize(dataset_dest_file(args, output_prefix, None, "idx"))
    return res


Sergey Edunov's avatar
Sergey Edunov committed
322
def dataset_dest_prefix(args, output_prefix, lang):
323
    base = "{}/{}".format(args.destdir, output_prefix)
324
325
326
327
328
329
330
    if lang is not None:
        lang_part = ".{}-{}.{}".format(args.source_lang, args.target_lang, lang)
    elif args.only_source:
        lang_part = ""
    else:
        lang_part = ".{}-{}".format(args.source_lang, args.target_lang)

331
    return "{}{}".format(base, lang_part)
Sergey Edunov's avatar
Sergey Edunov committed
332
333
334
335


def dataset_dest_file(args, output_prefix, lang, extension):
    base = dataset_dest_prefix(args, output_prefix, lang)
336
    return "{}.{}".format(base, extension)
Ruty Rinott's avatar
Ruty Rinott committed
337
338
339


def get_offsets(input_file, num_workers):
340
    return Binarizer.find_offsets(input_file, num_workers)
Ruty Rinott's avatar
Ruty Rinott committed
341
342


Myle Ott's avatar
Myle Ott committed
343
def cli_main():
344
    parser = options.get_preprocessing_parser()
Myle Ott's avatar
Myle Ott committed
345
346
    args = parser.parse_args()
    main(args)
Myle Ott's avatar
Myle Ott committed
347
348
349
350


if __name__ == "__main__":
    cli_main()